Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech

Page created by Clinton Haynes
 
CONTINUE READING
Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech
Atmos. Meas. Tech., 13, 1357–1371, 2020
https://doi.org/10.5194/amt-13-1357-2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Analysis of flow in complex terrain using
multi-Doppler lidar retrievals
Tyler M. Bell1,2 , Petra Klein1,2 , Norman Wildmann3 , and Robert Menke4
1 School of Meteorology, University of Oklahoma, Norman, OK, USA
2 Cooperative Institute for Mesoscale Meteorological Studies, University of Oklahoma, Norman, OK, USA
3 Deutsches Zentrum für Luft- und Raumfahrt e.V., Münchener Str. 20, 82234 Weßling, Germany
4 Technical University of Denmark – DTU Wind Energy, Fredriksborgvej 399, Building 118, 4000 Roskilde, Denmark

Correspondence: Tyler M. Bell (tyler.bell@ou.edu)

Received: 29 November 2018 – Discussion started: 18 January 2019
Revised: 7 February 2020 – Accepted: 13 February 2020 – Published: 24 March 2020

Abstract. Strategically placed Doppler lidars (DLs) offer         1   Introduction
insights into flow processes that are not observable with
meteorological towers. For this study we use intersecting
range height indicator (RHI) scans of scanning DLs to cre-        Scanning Doppler lidar (DL) systems have proven to be use-
ate four virtual towers. The measurements were performed          ful in many different sectors of atmospheric study. They
during the Perdigão experiment, which set out to study atmo-      have been used for boundary layer meteorology (Klein et al.,
spheric flows in complex terrain and to collect a high-quality    2015; Fernando et al., 2015), wind energy research (Banta
dataset for the validation of meso- and microscale models.        et al., 2015; Newman et al., 2016; Choukulkar et al., 2017),
Here we focus on a period of 6 weeks from 1 May 2017              and other various fields of study (Sathe and Mann, 2013;
through 15 June 2017. During this Intensive Observation Pe-       Bonin et al., 2017). DLs measure the radial velocity along
riod (IOP) data of six intersecting RHI scans are used to         a beam in a high spatial and temporal resolution. Different
calculate wind speeds at four virtual towers located along        scanning strategies can give different insights into the flow
the valley at Perdigão with a temporal resolution of 15 min.      field surrounding the DLs. For example, a plan position in-
While meteorological towers were only up to 100 m tall, the       dicator (PPI) scan gives a representation of the spatial vari-
virtual towers cover heights from 50 to 600 m above the val-      ability in the horizontal by scanning at a fixed elevation and
ley floor. Thus, they give additional insights into the complex   only moving in azimuth, while a range height indicator (RHI)
interactions between the flow inside the valley and higher up     scan essentially gives a cross section of the flow by staying
across the ridges. Along with the wind speed and direction,       at a fixed azimuth but changing elevation.
uncertainties of the virtual-tower retrieval were analyzed. A        By applying assumptions to the flow, one can use these dif-
case study of a nighttime stable boundary layer flow with         ferent scan strategies to derive the two-dimensional (2D) and
wave features in the valley is presented to illustrate the use-   three-dimensional (3D) wind from a single DL. The simplest
fulness of the virtual towers in analyzing the spatially com-     techniques to derive the wind speed and direction are the
plex flow over the ridges during the Perdigão campaign. This      velocity azimuth display (VAD) technique (Browning and
study shows that, despite having uncoordinated scans, the         Wexler, 1968) and the Doppler beam swinging (DBS) tech-
retrieved virtual towers add value in observing flow in and       nique (Strauch et al., 1984). However, these techniques make
above the valley. Additionally, the results show the virtual      the assumption that the wind is horizontally homogeneous in
towers can more accurately capture the flow in areas where        order to retrieve the wind speed and direction; this is often
the assumptions for more traditional DL scan strategies break     not the case in boundary layer meteorology and can introduce
down.                                                             errors into the wind estimates. One area where the assump-
                                                                  tion of horizontal homogeneity is likely invalid is in the study
                                                                  of complex terrain, where different beams of a scan pattern

Published by Copernicus Publications on behalf of the European Geosciences Union.
Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech
1358                              T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals

may sample different flow features. While in flat homoge-
neous terrain the accuracy of wind speed measurements with
conically scanning DLs was found to be within a few percent,
in complex terrain errors as large as 10 % are not uncommon
(Bingöl et al., 2009; Bradley et al., 2015). Correction tech-
niques such as Leosphere’s Flow Complexity Recognition
(FCR) algorithm focus on correcting errors introduced by
terrain by using simple flow models (see Leosphere, 2020).
Other techniques using a single DL have been developed to
limit the assumption of horizontal homogeneity (Waldteufel
and Corbin, 1979; Wang et al., 2015).
   In complex terrain, retrievals using multiple DLs with
beams that intersect in the same volume of air may be needed
to accurately measure the 3D wind field. By combining mul-
tiple different radial wind vectors, one can solve the 3D trans-
formation matrix to get the 3D wind vector without applying        Figure 1. Map of DLs, meteorological towers, and virtual towers
any assumptions to the flow. There are multiple ways that this     at the Perdigão site. The color fill corresponds to the height of the
can be done. For example, coplanar RHI scans were used in          terrain above sea level, red stars are virtual-tower locations, blue
the Terrain-induced Rotor Experiment to study rotors caused        triangles are DL locations, and green diamonds are instrumented
by mountains (Hill et al., 2010), multi-Doppler scan strate-       meteorological towers. Tower 9 is a 60 m tower and Tower 25 is a
gies were used in the Perdigão 2015 and 2017 experiments           100 m tower. CLAMPS was located at the Orange Grove site.
to measure wind turbine wake deficits in highly complex ter-
rain (Barthelmie et al., 2018; Menke et al., 2018; Wildmann
et al., 2018a, b), and virtual towers (VTs) were used in the       (Mann et al., 2017). As wind energy becomes more popular,
Joint Urban experiment in 2003 (Calhoun et al., 2006).             the need for models to accurately predict energy output from
   Multi-Doppler measurements can augment more tradi-              wind resources becomes increasingly important. NEWA pro-
tional observation strategies and are highly adaptable to an       vides a standard dataset of wind energy resources for Eu-
experiment’s science objectives, but they are not perfect.         rope through the creation of new and improved meso- and
Though some precision is lost due to volumetric averag-            microscale models and modeling techniques. One important
ing, the results are generally accurate to within 0.2 m s−1        aspect of developing these modeling techniques is verify-
(Damian et al., 2014; Pauscher et al., 2016; Debnath et al.,       ing their accuracy against high-quality datasets. Therefore,
2017). Higher levels of uncertainty have been found with           one of the main goals in the creation of the NEWA is devel-
more complicated scan strategies (Choukulkar et al., 2017).        oping new modeling methodologies which have been tested
For multi-Doppler retrievals, the magnitude of uncertainty         against observations from various measurement campaigns
was directly related to the ability to precisely coordinate        conducted in different landscapes and climates.
scans (Wildmann et al., 2018b). Additionally, higher levels           To validate numerical models, detailed measurements of
of uncertainty are present during unstable conditions (New-        the flow at multiple scales are required (Banta et al., 2013;
man et al., 2016).                                                 Fernando et al., 2015). To this day, data collected from the
   This study assesses how well traditional methods and 2D         Askervein Hill project in 1982 (Taylor and Teunissen, 1987)
and 3D multi-Doppler measurements perform in a com-                are still a standard dataset to validate models and to test
plex setting. Additionally, it provides an analysis of terrain-    how they handle flow over complex terrain. Some of the ex-
induced uncertainties compared to both multi-Doppler mea-          periments taking place under the NEWA are meant to aug-
surements and traditional single-Doppler wind estimation           ment the measurements from Askervein while adding an-
techniques. One challenge was however the fact that all of         other layer of complexity with observation campaigns fo-
the techniques applied have limitations, and a true reference      cusing on flow phenomena that current modeling techniques
dataset for quantifying the measurement errors did not exist.      are known to have difficulties with (e.g., vegetation canopies,
This is further addressed in Sect. 4.1.                            steep slopes, double-ridge configuration).
                                                                      The Perdigão experiment is one of the measurement cam-
                                                                   paigns that took place for NEWA (Mann et al., 2017).
2   The experiment                                                 Perdigão is a small municipality located in central Portugal
                                                                   that sits adjacent to two parallel ridges of nearly equal height
The data presented in this study were collected during the         with steep slopes separated by a valley called Vale do Co-
Perdigão field campaign during the spring and early sum-           brão. The ridge axes extend northwest to southeast and are
mer of 2017, which is one of multiple experiments conducted        approximately 1.4 km apart (Fig. 1). Previous wind-tunnel
in order to build the New European Wind Atlas (NEWA)               and numerical studies (e.g., Lee et al., 1987; Grubišić and

Atmos. Meas. Tech., 13, 1357–1371, 2020                                               www.atmos-meas-tech.net/13/1357/2020/
Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech
T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals                                     1359

Stiperski, 2009; Rapp and Manhart, 2011) have often focused          cated with the CLAMPS DL. The CU DL measures the wind
on the flow over sinusoidal hills. Two parallel ridges are the       speed and direction using the DBS technique (see Rhodes
best approximation in nature to study flow over periodic, si-        and Lundquist, 2013).
nusoidal type terrain.                                                  The German Aerospace Center (DLR) contributed three
   Additionally, there is a single 2 MW wind turbine located         Leosphere Windcube 200S scanning DLs upgraded with
on the southwest ridge. Wind resources are significantly af-         the Technical University of Denmark’s (DTU) WindScanner
fected by terrain and land cover. For example, winds at hub          software. Two of these DLs performed continuous RHI scans
height are much higher at the top of a hill than they are on the     in the cross-valley direction, which resulted in an RHI ap-
lee side of the hill. The flow on the lee side of the hill depends   proximately every 30 s. One of these DLs was located up on
largely on the stability conditions, wind speeds, etc. These         top of the NE ridge (DLR no. 1) and performed RHI scans
conditions are important when determining sites where wind           to the SW, capturing one horizontal component of the wind.
turbines will consistently be able to produce energy.                The other DL (DLR no. 2) was located on the slope of the
   Leading up to the Intense Observation Period (IOP) in the         NE ridge and performed RHIs to the SW as well.
spring of 2017, a meteorological tower had been operating               In addition to the DLR DLs, DTU operated eight DLs of
on the SW ridge for a few years. This was used to con-               the same kind on top of the ridges. Six of these were config-
struct a climatology of the wind directions over the ridge.          ured to do coplanar scans inside the valley so the horizontal
According to the climatology, winds often were found to be           wind in the plane and the vertical velocity could be retrieved.
directed perpendicular to the two ridges (Vasiljević et al.,        These DLs also operated in a continuous scan mode and pro-
2017). This provided a good opportunity to study the flow in         duced a new RHI every 24 s.
a double-ridge setting. To gather the desired observations, a           The RHIs from DLR no. 1 and DLR no. 2 overlapped
multinational group of scientists from Europe and the United         in a coplanar fashion, so by combining these DLs with the
States converged in Perdigão to measure complex flow at un-          CLAMPS DL scans, it is possible to retrieve the three-
precedented spatiotemporal scales. A combination of mete-            dimensional wind field in the form of a virtual tower where
orological towers, DLs, radiometers, and other remote and            the three planes intersect. Due to its positioning, DLR no. 2
in situ platforms were dispersed throughout the valley (see          was able to capture more of the vertical component of the
Fernando et al., 2019).                                              wind in the location of the virtual tower, which allowed the
                                                                     retrieval of the three-dimensional wind vector.
2.1   Virtual towers                                                    Regarding the DTU DLs, only the DLs from each coplanar
                                                                     cross section that reached deeper into the valley were used.
For this study, multiple different DL configurations were            This resulted in three possible 2D horizontal wind retrievals
jointly analyzed to retrieve the 3D wind vector in the val-          using WS2, WS5, and WS6. The 2D retrievals assume there
ley. In total, six DLs from three different institutions were        is no vertical velocity and thus only provide the along- and
combined to retrieve virtual towers multiple times per hour.         cross-valley wind components.
Figure 1 shows the location of the DLs considered in this               In total, four virtual towers distributed along the valley are
study. Details about the scanning strategies for each DL can         retrieved every 15 min when the CLAMPS DL performed its
be found in Table 1.                                                 along-valley RHI. The virtual towers typically cover heights
   The University of Oklahoma (OU) deployed the Col-                 from 50 to 600 m above the valley floor depending on the
laborative Lower Atmospheric Mobile Profiling System                 minimum and maximum height of RHI intersection. It should
(CLAMPS, Wagner et al., 2019) at the Orange Grove site,              be noted that different to previous virtual-tower retrievals
which is located inside the valley and served as central ob-         (Calhoun et al., 2006; Hill et al., 2010; Damian et al., 2014;
servation hub with a multitude of in situ and profiling instru-      Pauscher et al., 2016; Debnath et al., 2017) the DL data used
ments during the IOP (see Fig. 5 in Fernando et al., 2019).          in our study were not collected using coordinated scans that
CLAMPS includes a Halo Photonics scanning DL that per-               were specifically programmed for 2D and 3D wind retrievals.
formed both cross- (NE to SW) and along-valley (NW to SE)            Instead, our objective was to test the quality of virtual-tower
RHI scans every 15 min. In addition, a 70◦ PPI scan was per-         retrievals in complex flows using data from uncoordinated
formed every 15 min preceding the RHIs. The remainder of             scans with overlapping sampling volumes.
the time, the DL was in stare mode to get vertical-velocity
statistics. During Perdigão, the CLAMPS DL was config-
ured such that the first usable range gate was 75 m. CLAMPS          3   Methods
also utilizes an Atmospheric Emitted Radiance Interferome-
ter (AERI, Knuteson et al., 2004a, b) and a HATPRO mi-               When combining data from the uncoordinated DL scans in
crowave radiometer (Rose et al., 2005) for boundary layer            the virtual-tower retrievals, several analysis steps were nec-
temperature and humidity profiling (Wagner et al., 2019).            essary. First, data from each DL were converted to a com-
   The University of Colorado (CU) operated a Leosphere              mon coordinate system. Once the X–Y location of the RHI
V1 Windcube Profiling DL at the Orange Grove site, colo-             intersection (i.e., the virtual tower) was determined, mini-

www.atmos-meas-tech.net/13/1357/2020/                                                  Atmos. Meas. Tech., 13, 1357–1371, 2020
Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech
1360                               T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals

Table 1. Characteristics of the RHI scans performed by each DL used in this study and the virtual towers they were used for.

                   Lidar ID            Lidar      Azimuth   Minimum      Maximum         Update      Elevation            Virtual
                                     number                 elevation     elevation      time        ASL                  towers
                   OU CLAMPS             131         318◦       7.5◦             175◦    15 min      297 m              1, 2, 3, 4
                   DLR no. 1             172       236.1◦      −7.4◦            45.1◦    ∼ 30 s      454 m                       3
                   DLR no. 2             109       236.4◦       8.5◦             122◦    ∼ 30 s      316 m                       3
                   WS2                   102        54.7◦     −18.8◦            15.7◦    24 s        474 m                       2
                   WS5                   105        52.3◦     −12.8◦            21.7◦    24 s        476 m                       4
                   WS6                   106        42.2◦     −16.8◦            17.7◦    24 s        473 m                       1

mum and maximum heights were manually determined for                    3.1     Uncertainty analysis
each of the virtual towers. A vertical spacing of 10 m was
selected for the virtual towers to ensure that we had unique            Possible uncertainties contained in the retrieval are analyzed
DL range gates to derive the 3D/2D winds for each height.               using an idealized scheme derived from the methods dis-
Using the location of these points, the azimuth (θ̂i ) and el-          cussed in Hill et al. (2010):
evation (φ̂i ), relative to each DL, were calculated and used
to linearly interpolate the radial velocity from each DL to a                   "                            #1/2
                                                                                  N 
                                                                                      ∂ui 2
                                                                                 X       
                                                                                                       2
new radial velocity at the point of the tower (V̂ri ). We evalu-        ui =                     (ri ) ]          ; i = 1, 2, 3,    (2)
ated the difference of cubic spline interpolation to linear in-                   j =1
                                                                                         ∂rj
terpolation and found that differences of less than 0.1 m s−1           where ri is the possible random error associated with each
occur, which is well within the uncertainty we set for the DL           DL line-of-sight wind speed measurement, ui is the Carte-
radial wind speed measurement. Since neither cubic spline               sian velocity component of the wind, and rj is the radial ve-
nor linear interpolation can be assumed to be a true repre-             locity vector for each DL. This results in a formula for the er-
sentation of the atmospheric flow field, we decided to apply            ror for each velocity component given a combination of the
the simpler, computationally more efficient method of linear            various azimuths, elevations, and ranges for each lidar. For
interpolation. From there, retrieving the 3D/2D wind vector             this study, it is assumed that ri is 0.2 m s−1 based on prior
is as simple as solving the 3D transformation matrix for the            experience with the systems used (Kigle, 2017; Wildmann
radial velocity vector to get                                           et al., 2018a). Using a coordinate system that was aligned
                                                                        with the valley of the Perdigão site and choosing param-
                                                                        eters representative of the scan configurations of the vari-
                                                  #−1     
                                                                        ous DLs used, we computed the uncertainty of the retrieved
                                                      V̂r1
" # "
 u   sin θ1 cos φ1       cos θ1 cos φ1   sin φ1
 v = sin θ2 cos φ2       cos θ2 cos φ2   sin φ2      V̂r2 , (1)       along-valley, across-valley, and vertical-velocity components
 w   sin θ3 cos φ3       cos θ3 cos φ3   sin φ3       V̂r3              (Fig. 2). The results clearly show that with increasing height,
                                                                        as the DL beams point increasingly vertical, uncertainties in
                                                                        the horizontal wind components get larger and uncertainties
where u is the velocity in the east–west direction, v is the ve-        in the vertical velocity get smaller. It can further be noted that
locity in the north–south direction, w is the vertical velocity,        the errors in the horizontal wind components in the cross-
Vri is the interpolated radial velocity from the DL, φi repre-          valley direction are much larger in the 3D retrieval. This is
sents the elevation, and θi represents the azimuth angles.              due to the proximity of VT3 to the CLAMPS DL; gates at
   Temporal resolution of the virtual towers is limited to              much higher elevation angles must be used, meaning less of
the time resolution of the CLAMPS RHI. Since the scans                  the horizontal component of the wind in the along-valley di-
were not coordinated to sample the same volume of space                 rection is contained in the measured radial velocities.
simultaneously, a time window needed to be determined.
Choukulkar et al. (2017) used a time window of 15 s when                3.2     Impact of terrain on 2D towers
comparing uncoordinated multi-Doppler retrievals to a sonic
anemometer and found reasonable agreement. For this study,              Since we had only one site where all three components of the
all the time periods considered took place overnight, which             wind vector could be retrieved but had multiple 2D virtual-
generally had more steady flow due to the lack of turbulent             tower sites where the horizontal wind vectors were retrieved
mixing. Therefore, a time window of 60 s is used. While most            assuming that the vertical wind velocity is negligible, it is im-
of the data fell within this window, there are time periods             portant to quantify the errors introduced in the 2D retrievals.
late in the IOP where retrievals are not possible due to the            This is particularly important in complex terrain where large
CLAMPS RHI starting too late to be captured in the 60 s win-            vertical velocities are often present. To do this, a theoreti-
dow.                                                                    cal idealized setup was constructed (Fig. 3). DL1 and DL2

Atmos. Meas. Tech., 13, 1357–1371, 2020                                                        www.atmos-meas-tech.net/13/1357/2020/
Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech
T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals                                       1361

                                                                         the 2D towers retrievals are violated and contributions of the
                                                                         vertical motions to the radial velocities will be falsely pro-
                                                                         jected into horizontal motions causing errors in the along-
                                                                         and across-valley wind components. Multiple vertical veloc-
                                                                         ities were selected that mimic the range of the observed ver-
                                                                         tical stare data collected with the CLAMPS DL throughout
                                                                         the IOP, which allows us to get quasi-realistic assessments of
                                                                         the errors inherent in the 2D virtual-tower retrievals.
                                                                            Figure 4 shows expected errors in horizontal wind speed
                                                                         and direction that arise if the vertical component of the wind
                                                                         is assumed to be zero. The dashed vertical lines shown in
                                                                         Fig. 4 indicate the errors that can be expected for the vari-
                                                                         ous virtual towers at the Perdigão site given the location of
                                                                         the different DLs employed in the retrievals. It becomes im-
                                                                         mediately apparent that not accounting for vertical veloci-
Figure 2. Uncertainty from the various virtual towers. Solid lines       ties can introduce large errors in wind direction estimation at
are uncertainty in along-valley direction, dashed lines are uncer-       ridge height. Vertical velocities as large as 2 m s−1 were of-
tainty in the cross-valley direction, and dotted lines are uncertainty
                                                                         ten observed in the CLAMPS DL vertical stares, which can
in w. The color corresponds to the virtual tower (see Fig. 1). Note
                                                                         cause errors in wind direction estimation to be near 40◦ if
that most of the uncertainty is contained in the along-valley direc-
tion since the CLAMPS DL was scanning at a high elevation angle          the retrievals only use two DLs. Due to the close proximity
(elevations used ranged from 21 to 77◦ ). Additionally, the further      to CLAMPS, VT3 would have been the most erroneous 2D
away the virtual tower was from CLAMPS (i.e., the lower the ele-         retrieval. However, it was possible to retrieve the 3D winds
vation angle had to be), the lower the uncertainty.                      from this location.

                                                                         4     Results

                                                                         4.1    Selected case studies

                                                                         As mentioned in Sect. 1, one of the challenges of our study
                                                                         was the fact that none of the datasets available was free of
                                                                         limitations and all DL retrieval methods applied introduced
                                                                         measurement errors. In addition to the systematic uncertainty
Figure 3. Idealized setup to examine the errors associated with ne-      analyses presented in the previous section, we decided to il-
glecting the vertical velocity and only producing 2D virtual towers.     lustrate the strengths and weakness of the different retrieval
The setup is meant to closely mimic the positions of the CLAMPS          methods by selecting cases from the Perdigão IOP with flow
DL, DLR no. 1, and DLR no. 2 used in the 3D virtual tower. DL3 is        patterns of increasing complexity. Given the topography of
varied along x to simulate different elevation angles.                   the site, we assumed that the flow can be considered quasi-
                                                                         2D for wind directions perpendicular to the valley; i.e., un-
                                                                         der such conditions the flow variability along the valley is
are fixed in place, but DL3 is allowed to move in range                  expected to be small. Note that there can still be a compo-
away from the tower, thereby decreasing the elevation an-                nent of the wind directed in the along-valley direction; the
gle required to observe the same point. This setup is meant              distinction is that this component does not vary along the
to mimic the DL placements for the 3D tower (see VT3 in                  length of the valley contained inside the study domain (i.e.,
Fig. 1). For assessing the errors introduced by neglecting the           between WS5 and WS6). To ensure the flow was quasi-2D,
vertical wind velocity in the 2D virtual towers only DL1 and             the along-valley RHI scans from the CLAMPS DL were vi-
DL3 are then used for wind retrievals.                                   sually inspected to ensure any along-valley component of the
   Next, we selected different 3D wind speeds and directions             wind did not vary in the study domain.
at the intersection of all the DL beams and calculated the ra-              We focused the analysis on three cases: first we selected
dial velocities that are observed by each DL by rearranging              a case with quasi-2D flow and small vertical velocities, next
Eq. (1). By only using the radial velocities from DL3 and                we chose a quasi-2D flow case with large vertical velocities,
DL1, a 2D retrieval similar to the 2D retrievals done for the            and last we selected a fully 3D case with complex flow in-
real virtual towers can be simulated (with one DL on top of              teractions that are associated with strong variability of the
the ridge and with the CLAMPS system inside the valley).                 flow along the valley. The selected days from the Perdigão
However, for fully 3D flow fields, the assumptions used for              IOP were subjectively identified based on the analysis of

www.atmos-meas-tech.net/13/1357/2020/                                                      Atmos. Meas. Tech., 13, 1357–1371, 2020
Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech
1362                                 T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals

                                                                         be used to understand flow phenomena for a highly complex
                                                                         case.

                                                                         4.1.1   Quasi-2D flow with limited vertical motion

                                                                         For our simplest case, we selected a day with wind speeds
                                                                         that exceeded 7–10 m s−1 at or above ridge height with wind
                                                                         directions perpendicular to the ridge axis. This minimizes the
                                                                         spatial variability of the flow along the length of the val-
                                                                         ley. Additionally, we target a time period during which the
                                                                         observed vertical velocities were largely less than 0.5 m s−1
                                                                         by examining the vertical stare from the CLAMPS DL –
                                                                         12 June 2017 from 00:00 to 06:00 UTC fit these criteria.
                                                                            During this period, a shortwave trough was off the coast
                                                                         of Portugal. Winds at 500 mbar were approximately 10 m s−1
                                                                         from the SW. No discernible mesoscale surface features were
                                                                         present as a result of the trough. However, there are a num-
                                                                         ber of mesoscale circulations that dominate the flow around
                                                                         Perdigão (Fernando et al., 2019). Thermal flows from the
                                                                         Serra da Estrela to the north often compete against synoptic-
                                                                         scale flows. This can introduce unique layering of wind
                                                                         speeds and direction near the ground. Winds during this pe-
                                                                         riod were 5–10 m s−1 throughout the night (Fig. 5a) and were
                                                                         out of the northeast (Fig. 5d). A strong temperature inversion
                                                                         was present in the valley. Additionally, there was very little
                                                                         vertical motion and turbulence (Fig. 5g and j).
Figure 4. Wind speed and direction errors associated with ne-               By starting with a relatively simple case, it is possible to
glecting the vertical velocity in 2D virtual towers. Results are for     directly compare the retrieval methods with minimal viola-
R = 300 m and H = 200 as depicted in Fig. 3. These values were           tions to the underlying assumptions. Unfortunately, VT4 and
chosen to mimic virtual-tower measurements slightly above ridge
                                                                         VT1 did not meet the retrieval criteria because the CLAMPS
height. The line colors correspond to different vertical velocities
examined. The vertical lines correspond to the elevation for VT1,
                                                                         DL became slightly out of sync with WS5 and WS6, and
VT2, and VT3 for the given R and H ; thus these are the expected         the time difference between scans did not meet the retrieval
error for this point in space for these VTs. Note that as vertical ve-   criteria. However, for this application, having VT2 and VT3
locity increases, error in wind direction estimation increases even at   will suffice since they are the closest to each other spatially
relatively low elevation angles.                                         (Fig. 1). Differences between all the various systems com-
                                                                         pared to VT3 are shown in Fig. 6 with the various uncer-
                                                                         tainties propagated through the difference. The CLAMPS
                                                                         vertical stare indicates that VT3 is overestimating the verti-
DL scans and stability profiles. Additionally, data availability         cal velocity by approximately 0.5 m s−1 from −40 to 180 m
was taken into account.                                                  (Fig. 6c and f). In this layer, there are biases in the VT3 wind
   For the first case, we would expect that the VTs and                  speed and direction as well (Fig. 6d and e) while the other
single-Doppler retrievals agree well, even though the sites              instruments fall within their respective levels of uncertainty.
vary along the valley. We argue that in this case, the ob-               The wind speed of VT3 is biased by approximately 10◦ while
served differences between the retrieval methods are primar-             the wind speed is biased by approximately 0.5 m s−1 . Where
ily caused by measurement errors and uncertainties in the re-            the VT3 vertical velocities agree with the CLAMPS verti-
trieval methods and less by flow variability; i.e., the compari-         cal velocities above 180 m, the wind speeds and direction are
son allows us to assess the measurement errors introduced by             within the degree of uncertainty in the retrievals.
different retrieval approaches. The second case, with stronger
vertical motions while still being quasi-2D, allowed us to an-           4.1.2   Quasi-2D flow with strong vertical motion
alyze the amount of error vertical motion introduces into the
2D retrievals for a realistic scenario. The observed differ-             Similar to the previous case, we again targeted a time pe-
ences can then also be compared with the results from the                riod with wind speeds that exceeded 7–10 m s−1 at or above
systematic study described in Sect. 3.2 for cross-validation.            ridge height and wind directions that are perpendicular to the
Once the virtual-tower uncertainties are well characterized,             ridge axes, during which the flow along the ridges can be
it is then possible to illustrate how the virtual towers can             assumed to be more quasi-2D. As with the previous case,

Atmos. Meas. Tech., 13, 1357–1371, 2020                                                    www.atmos-meas-tech.net/13/1357/2020/
Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech
T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals                                    1363

Figure 5. Time–height plots of three nights observed by the CLAMPS system. The top row (a–c) shows wind speed derived from the 70◦
VAD scan, the second row (d–f) shows the wind directions from the VAD, the third row (g–i) shows 15 min averages of vertical velocity
derived from the CLAMPS vertical stare data, and the bottom row (j–l) shows the standard deviation from the averages. The left column
shows data from 8 May 2017, the middle column data from 12 June 2017, and the right column data from 14 June 2017. The height axis is
relative to the base of the wind turbine on the southwest ridge, which is a proxy for ridge height.

RHIs from all the DLs were inspected to ensure this was                 For this case, consistent differences between the 2D and
the case. However, this time we selected a time period with          3D retrievals can be noted where there are vertical veloci-
relatively strong vertical motions greater than 1 m s−1 . This       ties present in the profile. Based on Fig. 4, one would expect
allowed instantaneous double- and triple-Doppler virtual-            there to be approximately a 10 to 20◦ difference in wind di-
tower measurements to be contrasted to single-Doppler mea-           rection between the fully resolved 3D tower (VT3) and the
surements (such as VADs). Additionally, the error analy-             2D tower (VT2) at 100 m above ridge height, where the ver-
sis from Sect. 3.2 could be validated using the 3D tower.            tical velocity was approximately 1 m s−1 (Figs. 5h and 8c).
During the overnight hours of 14 June 2017, the short-               Figure 8e shows that the wind direction from VT2 differs by
wave trough mentioned in Sect. 4.1.1 had moved over the              10 to 15◦ from the wind direction at the CLAMPS VAD and
area. The valley was neutrally stratified through the night.         VT3, which lends credence to the previously discussed ide-
Winds after 03:00 UTC increased to approximately 10 m s−1            alized model and provides a simple way to estimate the un-
from the southwest (Fig. 5b and e), and a persistent wave            certainty in the 2D towers due to the vertical motion. There
formed, causing large and steady vertical velocities over the        are also slight differences in wind speed present. While VT3
CLAMPS site (Fig. 5h and k).                                         and the CLAMPS VAD agree within the range of uncertainty

www.atmos-meas-tech.net/13/1357/2020/                                                  Atmos. Meas. Tech., 13, 1357–1371, 2020
Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech
1364                                 T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals

Figure 6. Profiles of horizontal wind speed (a), wind direction (b), and vertical velocity (c) of the CU DBS (blue), CLAMPS VAD (orange),
VT2 (pink, 2D), VT3 (purple, 3D), Tower 25 (red), and Tower 9 (green) from 12 June 2017 at 02:55 UTC (discussed in Sect. 4.1.1).
Additionally, profiles of the difference relative to VT3 are shown (d–f). VT4 and VT1 are not available for this time period due to DL
malfunction and not meeting retrieval criteria, respectively. Note the height axis is relative to the base of the wind turbine, which is a proxy
for the height of the ridge. Additionally, the error bars on the CLAMPS DL stare profile (f) are the standard deviation of the vertical velocities
observed between 1 min before and 1 min after the virtual-tower time. Uncertainty is propagated through the difference, hence the appearance
of error bars on Tower 9, Tower 25, and the CLAMPS VAD in (d–f).

at ridge height and above, VT2 is consistently indicating                  than 3 m s−1 at ridge height (Fig. 5c and f). Wind directions
1 m s−1 higher wind speeds (Fig. 8a and d), which agrees                   vary widely since the mean wind is interacting with various
well with the expected offset in Fig. 4. Again, the time series            slope flows occurring inside the valley. Numerous periods of
of VT3 and VT2 fall within the range of wind speeds ob-                    strong vertical motion were observed over the CLAMPS site
served by the tower, even capturing the wind speed increase                due to waves forming as a result of the terrain. On 8 May, a
that occurred around 04:00 UTC (Fig. 7c and f).                            unique, double-wave feature was observed along with a recir-
                                                                           culation within the valley. This occurred during the period of
4.1.3   Spatially complex case                                             relatively low wind speeds around 05:00 UTC (Fig. 5c). The
                                                                           double-wave feature did not appear often during the rest of
While the two previous cases were selected with the intent                 the IOP and lends itself well as a good test for the virtual-
of assessing the retrieval methods by limiting the analysis to             tower retrievals, especially the 3D virtual tower, given its
quasi-2D flows, the final case was chosen to illustrate how                complex evolution over time and space.
useful the virtual towers can be for measuring the spatial                    As mentioned previously, the traditional DL technique for
variability of features along the valley. A moderately com-                retrieving the wind speed and direction with a single DL is
plex case, 8 May 2017, was selected. During this time, there               the VAD or the DBS technique. However, in the overnight
was a low-pressure system off the coast of the Iberian Penin-              hours of the Perdigão campaign, the assumption of horizontal
sula and a ridge over Perdigão. Aloft, winds over the IOP site             homogeneity is often violated due to flow phenomena created
were from the southwest at 15 m s−1 around 500 mbar.                       by the terrain. This is expected to be particularly prominent at
   The complexity of the flow for 7–8 May is shown in Fig. 5.              the lowest levels of the VAD or DBS profiles near the surface.
Generally, winds are from the northeast during the night of 7              We will now examine data from 8 May (Sect. 4.1.3) to assess
May. On 8 May, winds veered through the night and were less

Atmos. Meas. Tech., 13, 1357–1371, 2020                                                         www.atmos-meas-tech.net/13/1357/2020/
Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech
T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals                                        1365

Figure 7. Time series of wind speeds (a–c) and wind directions (d–f) from VT3 (orange points), VT2 (green points), and the 5 min averaged
data from Tower 25 (blue points) at 100 m for the three selected cases.

Figure 8. Same as Fig. 6 but for 14 June at 02:58 UTC when higher vertical velocities were observed.

www.atmos-meas-tech.net/13/1357/2020/                                                     Atmos. Meas. Tech., 13, 1357–1371, 2020
Analysis of flow in complex terrain using multi-Doppler lidar retrievals - Atmos. Meas. Tech
1366                                 T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals

Figure 9. Similar to Fig. 6, only with VT1 (brown) and VT4 (gray) included. Profiles are from 8 May at 00:32 UTC. In general, the methods
agree well above the ridge (generally greater than 100 m), but results begin to get noisy 100 m below ridge height. For the sake of readability,
only every third point has been plotted for the virtual towers.

the virtual towers in weak, highly variable flow. Using the               directions, the stare and the 3D virtual tower start to diverge.
spatially distributed virtual towers can be helpful to study              The differences could be due to each measurement represent-
these type of flows.                                                      ing a different time. The tower data are 5 min averages, while
   Figure 9 illustrates how single DL retrievals can break                the CLAMPS DL stare data and the CLAMPS/DLR virtual
down in complex terrain. A few things become apparent                     tower are instantaneous measurements at slightly different
from the comparison. In general, the CLAMPS VAD falls                     locations, so some differences are to be expected.
within the uncertainty of the VTs in wind speed and direc-                   Breakdowns in the single-Doppler DL retrievals can be ob-
tion greater than 100 m above the ridge. Looking closer at VT             served by closely examining Figs. 6 and 8, assuming VT3
retrievals above ridge height, there seems to be a consistent             is taken to be the most accurate representation of the wind
offset in wind direction between VT3 and each of the other                speed and direction. Similar to the highly complex case, there
2D towers (Fig. 9e), which can be explained by the vertical               is relatively good agreement well above ridge height, where
velocities, as discussed in the previous section. This offset             the flow is less affected by the terrain. Inside the valley how-
gets larger as the profile approaches ridge height, where the             ever, the flow is less horizontally homogeneous due to the
vertical velocities are larger.                                           complexity of the valley floor. This causes there to be dif-
   Around 100 m below ridge height, the different techniques              ferences between the virtual towers and the single-Doppler
start to diverge. Wind speeds stay relatively consistent be-              retrievals (for example, below −100 m in Fig. 8).
tween each technique below ridge height, but the directions                  In summary, Fig. 9 shows how all the towers (both virtual
are highly variable. In particular, the wind directions from the          and real) compare to one other. In general, the wind speeds
CU DBS scan differ greatly from VT3. The CLAMPS VAD                       all line up nicely. There are differences in the wind direction
tends to agree slightly more but still significantly differs from         though. The differences can likely be attributed to the error
the collocated VT3 just below ridge height.                               associated with the 3D towers discussed in Sects. 3.2 and
   The vertical velocities from VT3 and the CLAMPS DL                     4.1, though they could also be due to the complex flow field
vertical stare also agree within the uncertainty VT3 above                as well. As vertical velocities become larger at and slightly
ridge height. Around the same level as the wind speeds and                below ridge height, the spread in wind direction gets slightly

Atmos. Meas. Tech., 13, 1357–1371, 2020                                                       www.atmos-meas-tech.net/13/1357/2020/
T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals                                    1367

                                                                    speeds, periods with lower wind speeds tend to have more
                                                                    spatial heterogeneity. In order to visualize this heterogeneity,
                                                                    it is best to think of wind components in the cross-valley and
                                                                    along-valley sense.
                                                                        Analysis of the virtual towers coupled with the cross-
                                                                    valley RHIs from the DLR and DTU DLs often shows that
                                                                    the flow can not be considered entirely 2D, especially at
                                                                    lower wind speeds. In Fig. 11, cross-valley flow is approx-
                                                                    imately −4 m s−1 . No clear jet is present in the RHI or vir-
                                                                    tual towers, but there is a large recirculation in the valley.
                                                                    During this time period, flow was strong enough that the cir-
                                                                    culation was present in all the cross sections of the valley.
                                                                    However, the size and strength of the recirculation varied due
                                                                    to the topography of the valley floor. Where the recirculation
                                                                    is strongest, cross-valley wind speeds are slightly reduced in
                                                                    the virtual tower (VT2 and VT4). This is especially promi-
                                                                    nent in the lowest levels of the virtual towers near the cir-
                                                                    culation. This also appears in some of the levels well above
                                                                    the valley, which is significant because it implies that the re-
                                                                    circulation is correlated with changes in the flow well away
                                                                    from the feature itself. This recirculation could be a rotor
                                                                    or hydraulic jump type feature as observed in Neiman et al.
                                                                    (1988).
                                                                        Looking at the along-valley component of the flow, winds
                                                                    are very calm and near zero within the valley. There does
                                                                    appear to be a weak jet just above the ridge at 550 m. There is
                                                                    evidence of a wind speed maximum near the surface, which
                                                                    indicates there is a downslope flow. This feature was often
                                                                    observed overnight in the along-valley RHI from CLAMPS.
Figure 10. Scatter plots of the wind speeds (a) and direction (b)
                                                                    Though the virtual towers do not extend far enough into the
from VT2 (green) and VT3 (orange) vs. the 5 min averages of wind
speed and direction from Tower 25.                                  valley to capture this flow, the meteorological towers are able
                                                                    to capture the near-surface speed maximum associated with it
                                                                    (Fig. 9). This persistent flow adds to the three-dimensionality
larger. Additionally, the spread is approximately what would        of the flows present in the valley.
be expected from Fig. 4. However, there is still noticeable             Analysis of a different time period during the night of
spread in the lowest levels that is likely due to heterogene-       8 May further shows the complexity of the flows in the ter-
ity of the terrain. Additionally, Fig. 7 shows that VT3 agrees      rain. During this night, winds veered rather quickly aloft and
well throughout the night; however, VT2 contains a larger           weakened closer to the surface (Fig. 5). An interesting tran-
spread in wind directions. To determine if these are real, the      sition occurs and a unique double wave forms within the val-
flow needs to be examined in a spatial sense. Taking into con-      ley along with a shear layer (Fig. 12). Like the previous case,
sideration all the sources of uncertainty in the towers, it is      there is little along-valley component of the wind below the
possible to examine the flow in and around the valley at a          ridge. There is still a speed maximum in the down-valley flow
high level of detail. Overall, the virtual towers perform rea-      in the lowest 20–30 m of the CLAMPS RHI scan, which is
sonably well during each case study (Fig. 10). A slight posi-       indicative of the aforementioned downslope flow. The recir-
tive bias in wind direction was present in VT2 since it suffers     culation is also still present inside the valley and has a highly
from the errors described in previous sections.                     complex structure that differs based on the cross section of
                                                                    the valley. The recirculation looks to be largest in the VT2
4.2   Spatial analysis                                              cross section, similar to the previous case. This again appears
                                                                    to slow wind speeds upstream of the recirculation.
One of the main reasons Vale do Cobrão was chosen for
the experiment was the quasi-two-dimensional nature of the
ridges; they were thought to be the best way to represents          5   Conclusions
a series of periodic rolling hills and that flow perpendicu-
lar to the ridges could also be considered quasi-2D. While          Multi-Doppler analyses are a useful tool for understand-
this assumption is valid during some nights with higher wind        ing and quantifying wind characteristics in complex terrain.

www.atmos-meas-tech.net/13/1357/2020/                                                 Atmos. Meas. Tech., 13, 1357–1371, 2020
1368                                 T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals

Figure 11. (a) Along-valley cross section of the flow for 8 May at 00:32 UTC. The color fill is the radial velocity projected into the horizontal
from the CLAMPS RHI used to create the virtual towers. Positive values indicate flow in the +X direction, which is directed to the NW. The
overlaid vectors are the components of the virtual towers projected into the plane of the RHI scan, the inset plot shows how the terrain cross
sections are oriented and where they are located, and the dashed gray line is the height of the SW ridge. (b–e) Similar to the along-valley plot
on the left, only in the cross-valley sense. +X here points to the NE. The color of the line matches the cross section indicated in the terrain
inset. This plot corresponds to the profiles shown in Fig. 9.

Figure 12. Same as Fig. 11 but for 05:03 UTC. Note the different structure of the wave and the shear layers in the cross-valley winds. Also
note the different strengths of the recirculations.

Atmos. Meas. Tech., 13, 1357–1371, 2020                                                        www.atmos-meas-tech.net/13/1357/2020/
T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals                                        1369

Though the scans used for the virtual towers were uncoordi-        ley and gives a more complete picture of the spatial evolu-
nated, they can be useful for diagnosing flow conditions in        tion of features in complex terrain. They also allow a more
and above the valley. The virtual towers help fill the gap in      comprehensive validation dataset for numerical models in the
wind speed measurements inside the valley above the height         future. While simulated RHI scans from numerical models
of the physical towers (100 m) and where more traditional          can be compared to the observed RHIs, only the combination
DL scanning strategies may not be fully valid given the            of multiple DLs allows the retrieval of the 3D wind vector.
complexity of the site and in particular for this experiment       These data could one day be fed into a model-based wind re-
provide nicely distributed measurements along the valley at        trieval using advanced data assimilation methods to estimate
heights which are not captured by any other instrument.            the full 3D wind field in and around the valley to gain more
    Though the virtual towers are well suited to study the com-    insight into the governing physics of the flow.
plex flows observed during the IOP, they are not without lim-
itations. The uncertainty in the radial velocities needed to
be propagated through the retrieval. Due to the positioning        Data availability. Data of all instruments that were used in this
of the DLs used for the virtual towers, this meant that un-        study are stored on three mirrored servers owned by DTU, Univer-
certainty in the horizontal wind retrieval was larger with in-     sity of Porto, and the NCAR Earth Observing Laboratory (EOL),
creased height. However, vertical velocity retrievals on the       respectively. The data are publicly available through dedicated web
                                                                   portals of the University of Porto (https://perdigao.fe.up.pt, Univer-
single 3D virtual tower became more certain with height as a
                                                                   sity of Porto, 2020) and EOL (http://data.eol.ucar.edu/master_list/
larger component of the vertical velocity was observed in the      ?project=PERDIGAO, Earth Observing Laboratory, 2020). For this
radial velocities at the higher elevation angles. Additionally,    study, the datasets used are the CLAMPS DL data (Klein and Bell,
the 2D virtual towers made the assumption that there was no        2017), DTU DL data (Menke et al., 2019), and the NCAR 5 min
vertical velocity, which is often violated in this terrain. Due    tower data (UCAR/NCAR – Earth Observing Laboratory, 2019).
to this, they are prone to errors, particularly in wind direc-
tion. For example, with vertical velocities of 1 m s−1 , errors
of 20◦ were observed near ridge height. It was shown that          Author contributions. TMB, PK, NW, and RM were in charge of
these errors can be estimated and accounted for in an analy-       the conceptualization and methodology; TMB performed the formal
sis.                                                               analysis; PK, NW, and RM were in charge of the resources; TMB
    There are some uncertainties in the virtual towers that can    wrote the original draft; PK, NW, and RM reviewed and edited the
not be accounted for. Since the scans were uncoordinated,          original draft; TMB provided the visualizations; PK provided su-
the beams very rarely were observing the same point at the         pervision; PK and NW acquired funding.
same time. For this reason, cases with less steady flow may
have more errors. The cases chosen for this study were more
                                                                   Competing interests. The authors declare that they have no conflict
steady and quasi-2D than many nights during the campaign,
                                                                   of interest.
which made them ideal for testing the different profiling tech-
niques. Other differences may be caused by flow inhomo-
geneities that may occur along the valley. Though we tried to      Special issue statement. This article is part of the special
minimize this through careful case selection, the flow was         issue “Flow in complex terrain: the Perdigão campaigns
still complex, and the assumption of quasi-2D during the           (ACP/WES/AMT inter-journal SI)”. It is not associated with
cases presented in Sect. 4.1.1 and 4.1.2 may not be fully ful-     a conference.
filled. Additionally, making direct comparisons between the
different methodologies was made difficult by the lack of a
true reference measurement.                                        Acknowledgements. We would like to thank José Palma and José
    In terms of wind direction, despite the uncertainties in the   Caros Matos for their work to make this experiment a success. We
2D retrievals, the virtual towers agree better with the meteo-     would also like to thank Matt Carney and Edward Creegan for their
rological towers situated inside the valley than the VAD/DBS       hard work in getting CLAMPS running in Perdigão despite many
scans at those levels. However, in the analyzed cases wind         initial challenges, and we thank the entire Perdigão team for their
speeds at these levels were quite small, so a more detailed        collaborative spirit before, during, and after the campaign. Lastly,
                                                                   we appreciate the hospitality of the people in the municipality of
intercomparison between all the different methods of wind
                                                                   Alvaiade.
estimation is needed. This could be done with a DL simula-
tor that is able to use fine-scale model output from the valley
to mimic DL radial velocities and noise. These could then be
                                                                   Financial support. This research has been supported by the Divi-
fed into the virtual-tower retrieval and compared directly to      sion of Atmospheric and Geospace Sciences (grant no. 1565539)
the model output.                                                  and the Federal Ministry of Economy and Energy on the basis of
    Combining various types of profiler data with meteorolog-      a resolution of the German Bundestag under the contract numbers
ical towers on the ridge allows wind speeds to be sampled          0325518 and 0325936A.
as cross-valley flow enters, goes through, and exits the val-

www.atmos-meas-tech.net/13/1357/2020/                                                 Atmos. Meas. Tech., 13, 1357–1371, 2020
1370                                T. M. Bell et al.: Analysis of flow in complex terrain using multi-Doppler lidar retrievals

Review statement. This paper was edited by Jose Laginha Palma          Fernando, H. J. S., Pardyjak, E. R., Sabatino, S. D., Chow, F. K.,
and reviewed by two anonymous referees.                                  Wekker, S. F. J. D., Hoch, S. W., Hacker, J., Pace, J. C., Pratt,
                                                                         T., Pu, Z., Steenburgh, W. J., Whiteman, C. D., Wang, Y., Zajic,
                                                                         D., Balsley, B., Dimitrova, R., Emmitt, G. D., Higgins, C. W.,
                                                                         Hunt, J. C. R., Knievel, J. C., Lawrence, D., Liu, Y., Nadeau,
References                                                               D. F., Kit, E., Blomquist, B. W., Conry, P., Coppersmith, R. S.,
                                                                         Creegan, E., Felton, M., Grachev, A., Gunawardena, N., Hang,
Banta, R. M., Pichugina, Y. L., Kelley, N. D., Hardesty, R. M., and      C., Hocut, C. M., Huynh, G., Jeglum, M. E., Jensen, D., Ku-
  Brewer, W. A.: Wind Energy Meteorology: Insight into Wind              landaivelu, V., Lehner, M., Leo, L. S., Liberzon, D., Massey,
  Properties in the Turbine-Rotor Layer of the Atmosphere from           J. D., McEnerney, K., Pal, S., Price, T., Sghiatti, M., Silver, Z.,
  High-Resolution Doppler Lidar, B. Am. Meteorol. Soc., 94, 883–         Thompson, M., Zhang, H., and Zsedrovits, T.: The MATER-
  902, https://doi.org/10.1175/BAMS-D-11-00057.1, 2013.                  HORN: Unraveling the Intricacies of Mountain Weather, B. Am.
Banta, R. M., Pichugina, Y. L., Brewer, W. A., Lundquist, J. K.,         Meteorol. Soc., 96, 1945–1967, https://doi.org/10.1175/BAMS-
  Kelley, N. D., Sandberg, S. P., Alvarez II, R. J., Hardesty,           D-13-00131.1, 2015.
  R. M., and Weickmann, A. M.: 3D Volumetric Analysis of               Fernando, H. J. S., Mann, J., Palma, J. M. L. M., Lundquist, J. K.,
  Wind Turbine Wake Properties in the Atmosphere Using High-             Barthelmie, R. J., Belo-Pereira, M., Brown, W. O. J., Chow, F. K.,
  Resolution Doppler Lidar, J. Atmos. Ocean Tech., 32, 904–914,          Gerz, T., Hocut, C. M., Klein, P. M., Leo, L. S., Matos, J. C., On-
  https://doi.org/10.1175/JTECH-D-14-00078.1, 2015.                      cley, S. P., Pryor, S. C., Bariteau, L., Bell, T. M., Bodini, N.,
Barthelmie, R., Pryor, S., Wildmann, N., and Menke, R.: Wind             Carney, M. B., Courtney, M. S., Creegan, E. D., Dimitrova, R.,
  turbine wake characterization in complex terrain via inte-             Gomes, S., Hagen, M., Hyde, J. O., Kigle, S., Krishnamurthy,
  grated Doppler lidar data from the Perdigão experiment, J.             R., Lopes, J. C., Mazzaro, L., Neher, J. M. T., Menke, R., Mur-
  Phys. Conf. Ser., 1037, 052022, https://doi.org/10.1088/1742-          phy, P., Oswald, L., Otarola-Bustos, S., Pattantyus, A. K., Ro-
  6596/1037/5/052022, 2018.                                              drigues, C. V., Schady, A., Sirin, N., Spuler, S., Svensson, E.,
Bingöl, F., Mann, J., and Foussekis, D.: Conically scanning              Tomaszewski, J., Turner, D. D., van Veen, L., Vasiljević, N., Vas-
  lidar error in complex terrain, Meteorol. Z., 18, 189–195,             sallo, D., Voss, S., Wildmann, N., and Wang, Y.: The Perdigão:
  https://doi.org/10.1127/0941-2948/2009/0368, 2009.                     Peering into Microscale Details of Mountain Winds, B. Am.
Bonin, T. A., Choukulkar, A., Brewer, W. A., Sandberg, S. P., We-        Meteorol. Soc., 100, 799–819, https://doi.org/10.1175/BAMS-
  ickmann, A. M., Pichugina, Y. L., Banta, R. M., Oncley, S. P., and     D-17-0227.1, 2019.
  Wolfe, D. E.: Evaluation of turbulence measurement techniques        Grubišić, V. and Stiperski, I.: Lee-Wave Resonances over Double
  from a single Doppler lidar, Atmos. Meas. Tech., 10, 3021–3039,        Bell-Shaped Obstacles, J. Atmos. Sci., 66, 1205–1228, 2009.
  https://doi.org/10.5194/amt-10-3021-2017, 2017.                      Hill, M., Calhoun, R., Fernando, H. J. S., Wieser, A., Dörnbrack, A.,
Bradley, S., Strehz, A., and Emeis, S.: Remote sensing winds             Weissmann, M., Mayr, G., and Newsom, R.: Coplanar Doppler
  in complex terrain? a review, Meteorol. Z., 24, 547–555,               Lidar Retrieval of Rotors from T-REX, J. Atmos. Sci., 67, 713–
  https://doi.org/10.1127/metz/2015/0640, 2015.                          729, https://doi.org/10.1175/2009JAS3016.1, 2010.
Browning, K. A. and Wexler, R.: The Determination of Kine-             Kigle, S.: Wake Identification and Characterization of a Full Scale
  matic Properties of a Wind Field Using Doppler Radar,                  Wind Energy Converter in Complex Terrain with Scanning
  J. Appl. Meteorol., 7, 105–113, https://doi.org/10.1175/1520-          Doppler Wind Lidar Systems, Master’s thesis, Ludwig Maximil-
  0450(1968)0072.0.CO;2, 1968.                              ians Universität München, München, 2017.
Calhoun, R., Heap, R., Princevac, M., Newsom, R., Fer-                 Klein, P., Bonin, T. A., Newman, J. F., Turner, D. D., Chil-
  nando, H., and Ligon, D.: Virtual Towers Using Coher-                  son, P. B., Wainwright, C. E., Blumberg, W. G., Mishra,
  ent Doppler Lidar during the Joint Urban 2003 Disper-                  S., Carney, M., Jacobsen, E. P., Wharton, S., and Newsom,
  sion Experiment, J. Appl. Meteorol. Clim., 45, 1116–1126,              R. K.: LABLE: A Multi-Institutional, Student-Led, Atmospheric
  https://doi.org/10.1175/JAM2391.1, 2006.                               Boundary Layer Experiment, B. Am. Meteorol. Soc., 96, 1743–
Choukulkar, A., Brewer, W. A., Sandberg, S. P., Weickmann, A.,           1764, https://doi.org/10.1175/BAMS-D-13-00267.1, 2015.
  Bonin, T. A., Hardesty, R. M., Lundquist, J. K., Delgado, R.,        Klein, P. M. and Bell, T. M.: CLAMPS Scanning Doppler
  Iungo, G. V., Ashton, R., Debnath, M., Bianco, L., Wilczak, J.         Lidar Data, UCAR/NCAR – Earth Observing Laboratory,
  M., Oncley, S., and Wolfe, D.: Evaluation of single and mul-           https://doi.org/10.5065/D6HD7TDP, 2017.
  tiple Doppler lidar techniques to measure complex flow during        Knuteson, R. O., Revercomb, H. E., Best, F. A., Ciganovich, N. C.,
  the XPIA field campaign, Atmos. Meas. Tech., 10, 247–264,              Dedecker, R. G., Dirkx, T. P., Ellington, S. C., Feltz, W. F., Gar-
  https://doi.org/10.5194/amt-10-247-2017, 2017.                         cia, R. K., Howell, H. B., Smith, W. L., Short, J. F., and To-
Damian, T., Wieser, A., Träumner, K., Corsmeier, U., and                 bin, D. C.: Atmospheric Emitted Radiance Interferometer. Part
  Kottmeier, C.: Nocturnal Low-level Jet Evolution in a Broad Val-       I: Instrument Design, J. Atmos. Ocean Tech., 21, 1763–1776,
  ley Observed by Dual Doppler Lidar, Meteorol. Z., 23, 305–313,         https://doi.org/10.1175/JTECH-1662.1, 2004a.
  https://doi.org/10.1127/0941-2948/2014/0543, 2014.                   Knuteson, R. O., Revercomb, H. E., Best, F. A., Ciganovich, N. C.,
Debnath, M., Iungo, G. V., Brewer, W. A., Choukulkar, A.,                Dedecker, R. G., Dirkx, T. P., Ellington, S. C., Feltz, W. F., Gar-
  Delgado, R., Gunter, S., Lundquist, J. K., Schroeder, J. L.,           cia, R. K., Howell, H. B., Smith, W. L., Short, J. F., and Tobin,
  Wilczak, J. M., and Wolfe, D.: Assessment of virtual towers            D. C.: Atmospheric Emitted Radiance Interferometer. Part II: In-
  performed with scanning wind lidars and Ka-band radars dur-            strument Performance, J. Atmos. Ocean Tech., 21, 1777–1789,
  ing the XPIA experiment, Atmos. Meas. Tech., 10, 1215–1227,            https://doi.org/10.1175/JTECH-1663.1, 2004b.
  https://doi.org/10.5194/amt-10-1215-2017, 2017.

Atmos. Meas. Tech., 13, 1357–1371, 2020                                                   www.atmos-meas-tech.net/13/1357/2020/
You can also read